Why Does Testing Matter?
Counterfeit gold is a real and growing problem. Chinese tungsten-filled bars and coins have been documented by multiple dealers and assay offices since at least 2012. Tungsten is the counterfeiter’s material of choice because its density (19.25 g/cm3) is nearly identical to gold’s (19.32 g/cm3), making weight-based detection alone unreliable.
The seven methods below are ranked from least to most reliable. No single test is foolproof. Sophisticated fakes can pass three or four tests simultaneously. The goal is layered verification: stack multiple tests to raise the confidence level.
Method 1: Visual Inspection
Cost: Free Accuracy: Low What it catches: Obvious fakes, crude counterfeits
Examine the coin or bar under strong light with a loupe (10x magnification minimum). Look for:
- Mushy or poorly defined details, particularly on lettering and fine design elements
- Incorrect color. Pure gold (.999) is a warm, rich yellow. 22K gold (Eagles, Krugerrands) has a slightly deeper, more orange tone. If the color leans toward brass or is unusually bright, proceed with caution.
- Seams, filing marks, or irregular edges
- Weight that feels “off” in hand (gold is dense, and experienced handlers notice immediately when something is wrong)
Limitation: Visual inspection catches only low-quality fakes. Any competent counterfeiter produces coins that pass visual examination.
Method 2: Magnet Test
Cost: $5-15 (rare earth neodymium magnet required) Accuracy: Low to moderate What it catches: Steel, iron, and magnetic alloy fakes
Gold is diamagnetic, meaning it is very weakly repelled by a magnetic field. It should not attract to a magnet. Hold a strong neodymium magnet against the piece:
- Strong attraction: The piece contains iron or steel. Not gold.
- No reaction: Consistent with gold, but also consistent with copper, tungsten, lead, and other non-magnetic metals.
- Slight repulsion (slide test): Tilt the gold piece at 45 degrees and slide the magnet down the surface. On genuine gold, the magnet slides noticeably slower than on a non-metallic surface due to Lenz’s law eddy currents. This variant is more informative than simple attraction testing.
Limitation: Tungsten is not magnetic. The magnet test does not detect the most common sophisticated counterfeits.
Method 3: Ping Test (Acoustic Test)
Cost: Free (or $5 for a smartphone app like Bullion Test or CoinTrust) Accuracy: Moderate What it catches: Fakes with different internal density or composition
Gold has a specific resonant frequency when struck. A genuine 1 oz American Gold Eagle produces a sustained, high-pitched ring at approximately 6.09 kHz when balanced on a fingertip and tapped with another coin.
Smartphone apps record the ping and compare the frequency spectrum against known signatures for specific coins. The apps are surprisingly effective for standard bullion coins with well-documented acoustic profiles.
Limitation: The ping test works well for coins but is unreliable for bars (different geometries produce different frequencies). Filled coins with thin gold shells can produce frequencies close to genuine. The test is best used as one layer in a multi-test approach.
Method 4: Specific Gravity Test (Water Displacement)
Cost: $20-50 (precision scale needed, 0.01g resolution minimum) Accuracy: Moderate to high What it catches: Most base metal fakes, plated fakes
Gold’s specific gravity is 19.32 for pure gold and approximately 17.7-17.8 for 22K Eagle alloy. The test measures the ratio of the object’s weight in air versus its apparent weight when suspended in water.
Procedure:
- Weigh the piece in air. Record as W(air).
- Suspend the piece in distilled water using a thin thread or wire cradle, with the piece fully submerged and not touching the container.
- Record the apparent weight as W(water).
- Specific gravity = W(air) / (W(air) - W(water))
Expected values:
- .999 gold: 19.32
- .9167 gold (22K Eagle): 17.7-17.8
- .9999 gold (Maple Leaf): 19.32
- Tungsten: 19.25
- Lead: 11.34
- Copper: 8.96
Limitation: Tungsten’s density is close enough to gold that a solid tungsten fake passes this test. However, a tungsten core with a gold shell often creates detectable specific gravity deviations because the shell thickness varies. This test reliably catches lead, copper, and brass fakes.
Method 5: Acid Test
Cost: $15-30 (acid test kit) Accuracy: High for surface composition What it catches: Plated fakes, gold-filled items, wrong karat
Acid test kits use nitric acid and aqua regia solutions calibrated for different karat levels. A small scratch is made on a testing stone, and the appropriate acid is applied to the gold streak left on the stone.
- Genuine gold of the correct karat: no reaction, streak remains
- Lower karat than claimed: streak dissolves or changes color
- Base metal: streak dissolves immediately
Procedure for coins and bars: To avoid damaging the piece, scratch on an inconspicuous area or use the edge. The scratch must penetrate any plating to reach the core material.
Limitation: Acid tests only verify surface and near-surface composition. A thick gold plating over a tungsten core passes the acid test if the scratch does not penetrate to the core. On a coin with 0.5mm+ gold thickness, the acid test confirms gold without detecting a filled interior.
Method 6: Sigma Metalytics Precious Metal Verifier
Cost: $800-1,200 (device plus wands) Accuracy: Very high What it catches: Most fakes, including many tungsten counterfeits
The Sigma Metalytics Precious Metal Verifier measures electrical resistivity (conductivity) through the full thickness of the piece using electromagnetic fields. Gold has a specific resistivity of 2.44 microohm-cm that differs from virtually every other metal.
Key advantages:
- Tests through plastic capsules, packaging, and assay cards without opening them
- Measures the bulk interior, not just the surface
- Distinguishes between karat levels (22K vs 24K)
- Takes 2-3 seconds per test
- Different wand sizes for coins, bars, and large bars
Why it detects tungsten: Tungsten’s electrical resistivity (5.65 microohm-cm) is more than double that of gold. Even a tungsten core with a gold shell produces a reading outside the acceptable range for pure gold, because the electromagnetic field penetrates the shell and measures the composite conductivity.
Limitation: The Sigma is not infallible. Some sophisticated multi-layer composites can potentially fool it, though documented cases are extremely rare. The device also requires the correct wand size for the piece being tested, as using an undersized wand reduces penetration depth.
For dealers and serious collectors, the Sigma is the standard desktop verification tool. For individual investors buying a few coins per year, the cost is harder to justify unless peace of mind has significant personal value.
Method 7: XRF (X-Ray Fluorescence) Analysis
Cost: $15,000-50,000 (device), or $25-75 per test at a dealer/assay office Accuracy: Highest available for non-destructive testing What it catches: Essentially everything, with one caveat
XRF guns fire X-rays at the sample and analyze the fluorescent X-rays emitted back. The energy spectrum identifies exact elemental composition. A handheld Thermo Fisher Niton or Olympus Vanta provides precise percentages of gold, silver, copper, and any other elements present.
The caveat: Standard XRF penetrates only 10-50 microns into the sample surface. A thick gold plating over a different core is not detected by XRF alone. However, combining XRF (confirming surface composition) with a Sigma test (confirming bulk conductivity) or specific gravity test creates a verification stack that catches every known counterfeit type.
For individual investors, XRF testing is available at most coin dealers, pawn shops with precious metals operations, and independent assay offices. Call ahead and expect to pay $25-75 per item.
What Fakes Pass Which Tests?
| Fake Type | Visual | Magnet | Ping | Specific Gravity | Acid | Sigma | XRF |
|---|---|---|---|---|---|---|---|
| Brass/copper plated | Pass | Pass | Fail | Fail | Fail | Fail | Pass* |
| Lead core, gold plated | Pass | Pass | Fail | Fail | Pass** | Fail | Pass* |
| Tungsten core, gold shell | Pass | Pass | Fail*** | Pass**** | Pass** | Fail | Pass* |
| Solid tungsten, gold plated | Pass | Pass | Fail | Pass | Fail | Fail | Pass* |
| Steel core | Pass | Fail | Fail | Fail | Fail | Fail | Pass* |
*XRF reads surface only. **If scratch does not penetrate shell. ***Depends on shell thickness. ****If shell is uniform thickness.
The Sigma Metalytics tester is the only sub-$2,000 device that detects the most dangerous counterfeits: tungsten-cored, gold-shelled pieces. This is why it has become the industry standard for dealer verification.
What Is the Best Strategy for Individual Investors?
The most effective counterfeit protection is not a testing device. It is buying from reputable, established dealers who authenticate their own inventory. Dealers like APMEX, JM Bullion, and SD Bullion test every product they handle. Their reputation and return policies provide a guarantee that no home test can match.
For investors who buy from secondary sources (eBay, local sellers, estate sales, coin shows), layered testing is essential:
- Minimum: Magnet test plus specific gravity test ($25 total investment). Catches 80%+ of fakes.
- Better: Add a ping test app ($5) and acid test kit ($20). Catches 95%+ of fakes.
- Best (sub-$1,500): Sigma Metalytics Precious Metal Verifier. Catches 99%+ of fakes.
- Definitive: Sigma plus XRF. Catches everything documented to date.
For more on the counterfeit landscape and red flags when buying, see our complete counterfeit detection guide.
Frequently Asked Questions
Can a fake gold coin pass all tests?
No single documented counterfeit has passed all seven tests simultaneously. The combination of specific gravity, Sigma conductivity, and XRF analysis catches every known fake type. The risk is relying on only one or two tests, which sophisticated counterfeits can defeat.
Is the magnet test reliable for gold?
The magnet test is useful for eliminating obvious fakes (steel or iron-based) but provides no protection against the most common counterfeits, which use non-magnetic metals like tungsten, copper, or lead. Passing the magnet test should not be taken as confirmation of authenticity.
How much does a Sigma Metalytics tester cost?
The base unit with standard wand runs approximately $800-900. Additional wands for different product sizes cost $150-200 each. The complete setup for coins, small bars, and large bars totals roughly $1,000-1,200. For a dealer or active buyer handling tens of thousands in metals, this is a reasonable insurance cost.
Should I test gold bought from a major dealer?
Reputable online dealers (APMEX, JM Bullion, SD Bullion, Monument Metals) authenticate every product. Testing dealer-purchased gold is optional for peace of mind but not necessary for safety. The risk is in secondary market purchases: eBay, Craigslist, local sellers, and unknown dealers.
Does gold testing damage the coin or bar?
Only the acid test causes any mark, and only if you scratch the piece directly (using a test stone avoids this). The magnet, ping, specific gravity, Sigma, and XRF tests are completely non-destructive. Coins and bars can be tested and returned to their capsules or packaging with no evidence of testing.